zsdonghao / u-net-brain-tumor Goto Github PK

Is anyone understood the code???????
Please provide the sources how shall i understand the tensorlayer programming .

when using u_net_bn in model,i get the error "train_op = tf.train.AdamOptimizer(lr_v, beta1=beta1).minimize(loss, var_list=t_vars)",and "ValueError: No variables to optimize",what's the reason,how can i figure out ?

Thank you for sharing your code.
I have trained the model and getting the .npz file which contains weights of the trained model.
Can you share the code for testing the images?
Thank you.

Could you show me the performance of your method without augmentation? I would like to know how many percents do we have if we apply augmentation. Thanks

@zsdonghao I am getting the memory error like this, What is the solution for this error?

Traceback (most recent call last):
File "train.py", line 279, in
main(args.task)
File "train.py", line 78, in main
y_test = (y_test > 0).astype(int)
MemoryError

Would it be possible to run the training model on a Nvidia gtx 1060 6GB ram . Would the training time be longer than 18 hours in this case?

In 2015 dataset , survival_data .csv is not present can you please explain what is there ?

as to "b_images", there is a transpose in TRAINING while not in EVALUATION

Hello,

I just want to ask if you apply the algorithm on 3D data (nii file) or 2D data (slice-by-slice)?

Thanks.

In the U-net paper ,the loss is calculated with a weight in order to balance the class frequencies.But in your code file,I can not find this code . By the way ,I am a Chinese ,my English is poor and I am new to tensorlayer.

How we can view the brain MRI and finally the segmented image.

while trainig (python train.py --task=all) i got error

Traceback (most recent call last):
File "train.py", line 250, in
main(args.task)
File "train.py", line 65, in main
import prepare_data_with_valid as dataset
File "C:\delete\shared_f\braintumor\u-net-brain-tumor-master\prepare_data_with
_valid.py", line 351, in
X_train_input = np.asarray(X_train_input, dtype=np.float32)
File "C:\anaconda\lib\site-packages\numpy\core\numeric.py", line 492, in asarr
ay
return array(a, dtype, copy=False, order=order)
MemoryError

how can i solve this one ?
thanks in advance for helping :)

Hello, Ihave 3k images with T1 brain tumor, and i manage to obtain 50% mean IOU. Can you put the pretrained weights, to see if transfer learning is helping?

I can not find training data in http://braintumorsegmentation.org/.Can anyone send me a download link? Thanks~，and The registered link is closed

while i am trying to run train.py file i get attribute error : tuple object has no attribute lower , i am using cpu for training ,

Hello Sir,

how can we visualize the feature in the network?
Can you give us some advices regarding that?

Thanks
Sami

How many numbers of images required to train Unet? for your task how much data you provided to the network without segmentation?

It reported that load checkpoint process was failed.
[TL] [!] Load checkpoint/u_net_all.npz failed!
Do you know how to fix it? Thank you.

while i was running this code this appears [!] Load checkpoint/u_net_all.npz failed! . what is the problem. i am running this code on cpu.

I got dice equal to one every epoch?
How to handle that?

Sorry to bother you,I tried to run the programme,I run 'train.py',but I got this error,but I don't know how to resolve it,thank you for your reply.

[TL] got 45: u_net/uconv1/bias:0 (1,)
INFO:tensorlayer: got 45: u_net/uconv1/bias:0 (1,)
[TL] ERROR: file checkpoint/u_net_all.npz doesn't exist.
ERROR:tensorlayer:ERROR: file checkpoint/u_net_all.npz doesn't exist.

I read you book and tried this demo
but I watch nvidia-smi ,it seems my GPU didn't work and my cpu are almost 100% usage
thank you !

Do you have implementation of brain tumor segmentation code directly in tensorflow without using tensorlayer? If yes, can you share the same?
Thank you.

there is a survival_csv_path, is that none, or have some data in it. ps: i train the model with 2015 data. is there any problem?

I have acquired the dataset which is BRATS 2015. It contains .mha files but the code deals with .nii.gz files. Need help in this since ita related to my thesis project.

I am running run.py but gives error:

(base) G:>cd BraTS_2018_U-Net-master

(base) G:\BraTS_2018_U-Net-master>run.py
[] creates checkpoint ...
[] creates samples/all ...
finished Brats18_2013_24_1
2019-06-15 22:05:45.959220: I tensorflow/core/platform/cpu_feature_guard.cc:141]
Your CPU supports instructions that this TensorFlow binary was not compiled to
use: AVX2
Traceback (most recent call last):
File "G:\BraTS_2018_U-Net-master\run.py", line 154, in

File "G:\BraTS_2018_U-Net-master\run.py", line 117, in main
t_seg = tf.placeholder('float32', [1, nw, nh, 1], name='target_segment')
NameError: name 'model' is not defined

Is anyone gone through the Evaluation ??

Hello
Can you tell me what the dice scores you can get on the brats17 verification set for your predecessor?

The registration time for Brat seems to have ended. Could you please provide an alternative for downloading the data?

Traceback (most recent call last):
File "C:\Users\zlshu\AppData\Roaming\Python\Python36\site-packages\numpy\lib\stride_tricks.py", line 251, in
if all(array.shape == shape for array in args):
SystemError: error return without exception set

if we select the DATA_SET=='small' and print (dev_index_HGG, test_index_HGG ), it returns the blank arrays.

I was wondering if you would share a pre-trained model.
I would need to run inference-only, and training the model is taking longer than expected.

Thanks for sharing this project!

Can you give me the survival_data.csv file for researching purpose?

• After conv: Tensor("u_net/conv8/leaky_relu:0", shape=(5, 1, 1, 512), dtype=float32, device=/device:CPU:0)
  Traceback (most re
  
  cent call last):
  File "train.py", line 250, in
  main(args.task)
  File "train.py", line 121, in main
  net = model.u_net_bn(t_image, is_train=True, reuse=False, n_out=1)
  File "/home/achi/project/u-net-brain-tumor-master/model.py", line 179, in u_net_bn
  padding=pad, act=None, batch_size=batch_size, W_init=w_init, b_init=b_init, name='deconv7')
  File "/home/achi/anaconda3/lib/python3.6/site-packages/tensorlayer/decorators/deprecated_alias.py", line 24, in wrapper
  return f(*args, **kwargs)
  TypeError: init() got an unexpected keyword argument 'out_size'

When I removed parts of the code that deal with LGG, an index error comes up.

I'm only interested in the HGG half of the dataset. Do you have an idea on how to solve this issue?

I used tensorflow-gpu1.8 and tensorlayer1.11, and I changed the original up3 = DeConv2d(conv4, 256, (3, 3), (nx/4, ny/4), (2, 2), name='deconv3') to up3 = DeConv2d(conv4, 256, (3, 3), strides=(2, 2), name='deconv3') in model.py and the lr=0.00001 .
According to readme, I restarted to train the model, but the 1-dice is always equal to one. Can you give me some advice to make it converge to near-perfect accuracy?

I have registered and got the login details. I have downloaded the dataset using details but I didn't find .csv file in it. Could you tell me, where the file is present.

Please join Slack, the invite link as follow:

https://join.slack.com/t/tensorlayer/shared_invite/MjExMjMzODAwNjI0LTE0OTk4ODEzMjUtNmVjNjQ4ZjMzMw

I can not find training data in http://braintumorsegmentation.org/.Can anyone send me a download link?
Thanks~

I do understand what is the survival_data.csv file, could u provide me?

I am trying to train the model using this command:
python train.py --task=all

But after the prepare_data is done it shows an error like this:

My version of tensorflow is 1.13.1. How can I solve this problem without downgrading the version of tensorflow?

How can I modify the code to save every slice of the segmented tumors?

how to get the training dataset ?

Please can someone send me the testing code !!!

is there a way to convert this UNet implementation to 3D?

hi

Can you help us where we can find csv data for 2015 data